Fig. 3

Trade-off between accuracy and coverage depends on a combination of methodological factors. The workflow with the highest coverage in this study could recall 75% of the original mock community members using mothur to process reads from the V4-V4 region of the 16S rRNA marker gene amplified using 1-step PCR. The workflows with the highest accuracy could only recall approximately 52% of the original mock community members using DADA2 or QIIME2. If high quality data was provided, the accuracy of mothur processed data was as good as data processed with DADA2 or QIIME2